SCI. MAR., 66 (2): 167-173 SCIENTIA MARINA 2002

Age, growth and longevity of the gray , capriscus (: Balistidae), from the Southeastern Brazilian Coast*

ROBERTO A. BERNARDES Laboratório de Ictiofauna do Instituto Oceanográfico, Universidade de São Paulo, Praça do Oceanográfico, 191-05508-900, São Paulo, Brazil. E-mail: [email protected]

SUMMARY: Age, growth and longevity of gray triggerfish Balistes capriscus from the coast near São Paulo were estimat- ed from first dorsal spine sections of 1,800 . The translucent zone was formed during winter (June, July and August) and the reproductive period (December, January, February). The von Bertalanffy growth equations were: FL = 514.9 [1 – e - 0.2625 (t + 0.0391) ] for males, and FL = 504.6 [1 – e - 0.2748 (t - 0.0304) ] for females. The longevity estimated was 11 years old for males and females. The instantaneous mortality rates estimated were 0.26 for males and 0.27 for females. The weight-length rela- tionships for both sexes of gray triggerfish together were Wt = 0.000004 FL 3.299.

Key words: gray triggerfish, Balistes capriscus, age, growth, longevity, Brazil.

INTRODUCTION 2,742.20 tons in 1992 and 650.59 tons in 1998 (Cas- tro, 2000). High rates of exploitation of Atlantic croaker This species occurs in the tropical and temperate (Micropogonias furnieri) and weakfish (Cynoscion zones of the Atlantic, from Nova Scotia (Canada) to jamaicensis) along the southeastern coast of Brazil Argentina, as well as along the African coast. In by the commercial fleet has led to a demand Brazil, it is particularly abundant along the coasts of for underutilised species. One of the underutilised Espírito Santo (20ºS), Rio de Janeiro (22ºS) and São species is the gray triggerfish, Balistes capriscus Paulo (24ºS). (Tetraodontiformes: Balistidae), a demersal species Information on age and growth of Balistes exploited in the state of São Paulo since the 1960s. capriscus using the first dorsal spine is available For many years it was included in the miscellaneous only from the southern coast of Africa (Caverivière fish category, but in the last 15 years, it has become et al., 1981; Ofori-Danson, 1989) and from the a very important resource and is currently northeastern Gulf of Mexico (Johnson and Saloman, one of the most commonly-caught commercial 1984), the later also including mortality estimates. species: 4,836.26 tons in 1987 (Bernardes, 1988), This paper presents the results of an investigation of age and growth of the gray triggerfish from the *Received June 9, 2000. Accepted February 23, 2001. south coast of the State of São Paulo in Brazil.

AGE, GROWTH AND LONGEVITY OF BALISTES CAPRISCUS 167 FIG. 1. – Map of São Paulo State south coast showing B. capriscus fishing area .

MATERIALS AND METHODS Whole vertebrae (from 526 fish) were observed under reflected light and the number of zones and Monthly samples (from December 1984 to type of growth of margin (translucent/opaque) were December 1985, except October 1985) were also recorded. The results of these analyses were obtained from commercial pair trawlers operating used to compare with those obtained from first dor- in waters between 24°S and 25°15’S (Fig. 1). A sal spine sections. total of 6,396 individuals were measured (fork , which are frequently used for aging length to the nearest millimeter) and 2,582 fish fish, were not used because in B. capriscus they are were weighed to the nearest 0.1 grams and sexed; very thin, fragile and small (1 to 2 mm). They are the gonads were weighed and the reproductive difficult to remove, keep and manage, so I avoided condition was assessed. For age determination the using them in this study. first dorsal spine and the first three vertebras were removed, cleaned with water and stored for later analysis. Transverse sections (180 µm in thickness) of 1,800 individuals were taken directly (without embedding in a polyester resin) from the first dorsal spines above the condyle with a low speed power saw (Fig. 2). the sections were mounted on glass slides and observed in transmitted light using a closed-circuit television (24x magnification). Translucent (light) bands on the spine sections were counted. The distance between the core and the pos- terior distal edge (spine radius = Rt) and the distance from the core to the distal edge of each translucent band (Ri) were measured (Fig. 3). The relationship between Rt and fork length (FL) was determined by the least squares method. The spine sections were analysed three times by two independent readers and FIG. 2. – Posterior view of first dorsal spine of Balistes capriscus the results were compared. showing the condyle and the level of sectioning.

168 R.A. BERNARDES FIG. 3. – Section of gray triggerfish first dorsal spine. A, spime section from a 2-year-old fish, collected in January 1985 (hyaline edge); B, spine section from a 2.2-year-old fish, collected in March 1985 (opaque edge); C, spine section from a 2.4-year-old fish, collected in May 1985 (opaque edge)

The periodicity of the growth zone formation RESULTS was estimated through the monthly frequency of translucent edges. The validation of age was based Length composition on the analysis of marginal increment, that is, the distance separating the last translucent growth mark The fork length (FL) of 6,396 fish landed by the from the edge of the spine (Cadwallader, 1978). commercial fleet ranged from 111 mm to 410 mm. The size of each fish was back-calculated for all The catch was composed mainly of fish between ages on dorsal spine sections using the Frasser-Lee 160 to 230 mm FL. There were very few specimens method (Bagenal and Tesch, 1978). The back-calcu- that measured less than 150 mm (N=52) and more lated mean length at age was used to calculate the than 350 mm (N=24). The mean fork length growth curve and to determine the growth parame- observed for 814 males was 216.4 mm (153 mm < ters of the Bertalanffy model (1938). The parameters FL< 347 mm) and for 1,240 females it was 203.5 of this growth equation (k, L∞ and t0) were estimat- mm (139 mm < FL< 334 mm). ed as suggested in Walford (1946) and Ford (1933). Length-frequency distribution (Fig. 4) showed Estimates of longevity and instantaneous natural basically two distinct groups from December 1984 mortality rate were calculated by Taylor’s equation to December 1985: the first group consisted of pre- (Taylor, 1960) as follows: adults (<200 mm FL) in the second year of life, and the second group was formed by older aged

A(0.95) = [2.9957/k] + t0 , mainly 2 to 4 years. About 46% of the catches were composed by immature fishes, a fork length at first for the longevity, and maturity being 169 mm for females and 200 mm for males (Bernardes and Dias, 2000). The smallest

M = - [ln (1 - P) /A(0.95)], specimens (110-130 mm FL) appeared from January to March, possibly indicating the benthic recruit- for instantaneous mortality, where ment of juveniles in this period.

P = FL/ L∞ = 0.95. Age estimation

The relationship between total body weight and The first dorsal spine transverse sections fork length for males, females and both sexes pooled showed a distinct translucent-opaque banding pat- were estimated. tern of deposition. The growth zones were distin-

AGE, GROWTH AND LONGEVITY OF BALISTES CAPRISCUS 169 FIG. 4. – Length-frequency histograms of Balistes capriscus caught by commercial fleet at São Paulo coast, between December 1984 and December 1985 (percentage of number). guishable in the upper anterior portion of the spine core and the first zone was not very distinct. The section. The core is small and located slightly interpretation of this portion is quite difficult due to upward in relation to the center of the spine (Fig. the fact that it corresponds to the first year of 3). The narrow translucent marks (annuli) are slow growth and is probably highly influenced by envi- growth zones and the wide opaque bands corre- ronmental conditions and food availability during spond to fast growth zone. The portion between the this stage.

170 R.A. BERNARDES narrow zone represents the first winter, the second narrow zone represents biological recruitment (changing of habitat and diet = 1 year old), the third narrow zone represents the second winter and the fourth narrow zone represents spawning period (2 years old). The largest fish aged (347 mm FL, cap- tured in July-85) presented 9 annuli on both the first dorsal spine and vertebrae, corresponding to 4,5 years. The smallest fish aged (139 mm FL, captured in February-85) presented 2 annuli, corresponding to 1 year of life. FIG. 5. – Marginal increment by month for Balistes capriscus captured off São Paulo (N = 274). Growth and longevity

A positive relationship was found between fork

length (FL) and radius of dorsal spine (Rt) and it was represented by the following equation:

2 FL = -26.9350 + 101.8125 Rt (r = 0.88; n = 363), for males; 2 FL = - 26.2179 + 100.8140 Rt (r = 0.89; n = 369), for females.

The mean back-calculated fork lengths at age of males and females are presented in Table 1.

FIG. 6. – Seasonal variation of percentage of hialine margin on dorsal The von Bertalanffy growth parameters varied spine and vertebrae of Balistes capriscus, captured off São Paulo. slightly between males, females and for all fish com- bined. The von Bertalanffy equations were: Of the 1,800 sections examined 83.5% were readable. The agreement between the two readers on males : FL = 514.9 [1 – e - 0.2625 (t + 0.0391) ] ; the spines was 78.6%. The oldest fish were the most females : FL = 504.6 [1 – e - 0.2748 (t - 0.0304) ] ; difficult to interpret. The analysis of the marginal all fish : FL = 509.5 [1 – e - 0.2687 (t + 0.0049) ] . increments by month for fish with 3 annuli on dor- sal spine has shown that the annuli are laid down The growth curves for gray triggerfish male and twice a year, during the summer (December to Feb- female are presented in Figure 7. ruary), and winter (June to August) (Fig. 5). These The weight-length relationship was represented results confirm that Balistes capriscus has two by the following equations: growth cycles per year. 3.265 2 The time of translucent and opaque zones forma- males: Wt = 0.000005 FL , (r = 0.97; n = 814); 3.340 2 tion was confirmed by observation of the marginal females: Wt = 0.000003 FL , (r = 0.96; n = 1,240); 3.299 2 zone type of dorsal spine and vertebrae over the year both: Wt = 0.000004 FL , (r = 0.96; n = 2,054). (Fig. 6). The translucent zone formation occurred during winter (June to August) and summer (December to March) which corresponds to the TABLE 1. – Back-calculated fork lengths (mm) of males and females of Balistes capriscus, from the southeastern coast of Brazil; n: spawning season. The opaque zone formation number of fish specimes; σ: standard deviation. occurred during the months of April and May, and from September to November. age male n σ female n σ both sexes January first was considered the fish birthday in order to attribute an annual scale of time to the 1 93.4 347 15.1485 93.5 211 13.6259 93.5 2 178 291 18.9621 180.1 182 21.6300 179.1 growth zones. Thus, a fish captured in January 3 247.4 68 23.0023 246.6 46 24.7012 247 showing four narrow and four wide zones on first 4 310.4 9 6.6008 312.7 7 14.6154 311.5 dorsal spine and on vertebrae is 2 years old. The first 5 346.5 4 0.5 347 2 - 346.7

AGE, GROWTH AND LONGEVITY OF BALISTES CAPRISCUS 171 ent environments, methods of capture or higher fish- ing mortality rates. Table 2 shows the results of dif- ferent authors. The period of fast growth for gray triggerfish from São Paulo coast, reflected by the deposition of an opaque zone, corresponds to the months of April and May and the period of September to November. On the other hand, the translucent zone, corresponding to the period of slow growth, appears from June to August (winter) and from December to February (summer). The factors that FIG. 7. – Growth curve of Balistes capriscus (male and female) from might be involved in translucent ring formation on São Paulo coast, based on back-calculated fork length. the structures of Balistes capriscus in this case are a decrease in temperature and food supply during The t-test indicated a significant difference (t = winter and the reproduction during summer. For 8.797; P< 0.05) between the weight-length relation- the African Balistes capriscus the formation of a ship for males and females, males being heavier than translucent zone in the summer is approximately females. equivalent to the beginning of gonadal maturity The longevity estimated for Balistes capriscus and the decrease in growth rates would be linked was 11 years old for males and females and the nat- to the spawning period (Caverivière et al., 1981). ural mortality rate (M) was 0.26 for males and 0.27 The African fish have a seasonal offshore migra- for females. tion to avoid cold coastal water, which is the result of upwelling (Anonymous, 1980, In: Johnson and Saloman, 1984). DISCUSSION It seems that the first translucent zone of our Bal- istes capriscus from the Brazilian coast, is deposited The age and growth of the gray triggerfish have in the first winter (0.5 years old, about 60-70 mm been reported from southwestern Africa by FL) and the second translucent zone is deposited Caveriviére et al. (1981) and by Ofori-Danson during summer (1 year old, 90-120 mm FL) due to (1989), and from the northeastern Gulf of Mexico changes of diet and habitat (pelagic to demersal). by Johnson and Saloman (1984). Johnson and Dooley (1972, In: Johnson and Saloman, 1984) Saloman (1984) determined that the translucent reported that the gray triggerfish is planktonic and zone formation of Balistes capriscus from the associated with during the first year of Gulf of Mexico, occurred during spring and sum- life, and after this period inhabitis the benthic area mer (from April to October with a peak during between 12 and 42 m in depth (Smith, 1976, In: June-July). In Senegal and the Ivory Coast, Johnson and Saloman, 1984). Caverivière et al. (1981) stated that the translucent Balistes capriscus was considered a moderate- zone was only visible after the second year of life. ly long-lived species by Johnson and Saloman They suggested two hypotheses regarding zone formation: A) one mark per year, or B) two marks TABLE 2. – Mean fork length (mm) at age and von Bertalanffy growth parameters (k, L∞ and t0) for Balistes capriscus,estimated per year. The fork lengths at age for the Gulf of by (1) this study; (2) Caverivière et al., 1981; (3) Ofori-Danson, Mexico fishes were larger after the first year than 1989; (4) Johnson and Saloman, 1984. predicted by both of hypothesis for Senegal fish. The fork lengths at age for Ivory Coast fish were São Paulo Senegal Ghana Gulf of Mexico (1) (2) (3) (4) smaller than Senegal and the Gulf of Mexico fish- Age Fork length Fork length Fork length Fork length es using the hypothesis of one zone formed per year. I 94.8 90 144 124.1 II 182 170 230 232.7 Comparing to data from other localities, Balistes III 250.8 238 292 306.8 capriscus from our coast were smaller in length at IV 305 290 330 357.3 V 347.7 324 391.8 age than those from the Ghana coast and Gulf of k 0.23 0.3058 0.43 0.38 Mexico and larger than those from the Senegal L∞ 506 407 408 466 t -0.13 -3.4 — -0.189 coast. These differences may be the result of differ- 0

172 R.A. BERNARDES (1984). These authors estimated that the oldest REFERENCES male gray triggerfish from the Gulf of Mexico was Bagenal, T.B. and F.W. Tesch. – 1978. Age and growth. In: T. 13 years old at 544 mm FL and the oldest female Bagenal (ed.) Methods for Assessment of Fish Production in was 12 years old (561 mm FL). Our estimates Fresh Waters, pp. 101-136. 3rd. Edition. Blackwell. BPI Hand book, 3. Oxford. were similar (11 years old for males and females). Bernardes, R.A. – 1988. Idade, Crescimento, Reprodução e Pesca However, our results indicated differences in do Peixe-Porco, Balistes capriscus (Gmelin,1788) capturado na costa sul do Estado de São Paulo (Brasil). M.Sc. thesis, asymptotic fork lengths (514.9 mm for males and Instituto Oceanográfico da Universidade de São Paulo, (Brasil). 504.6 mm for females) compared to those from the Bernardes, R.A. and J.F. Dias. – 2000. Aspectos da reprodução do peixe-porco, Balistes capriscus (Gmelin) (Actinopetrygii, Gulf of Mexico. I observed lower growth rates (k Tetraodontiformes, Balistidae) coletado na costa sul do Estado = 0.27) than those obtained from the Gulf of Mex- de São Paulo, Brasil. Revta. Bras. Zool., 17(3): 687-696. Bertalanffy, L. (von). – 1938. A quantitative theory of organic ico (0.38) and Africa (0.30). The differences growth (inquiries on growth laws. II). Hum. Biol. 10:181-213. between the estimated and reported k-values might Cadwallader, P.L. – 1978. Age, growth, and condition of the com- mon river Galaxias, Galaxias vulgaris Stolkell, in the Glentni be due to different environmental conditions River, Canterbury, New Zeland. Fish. Res. Bull., 17: 15-20. affecting growth. The analysis of the W -FL rela- Castro, P.M.G. de. – 2000. Estrutura e Dinâmica da Frota de Par- t elhas do Estado de São Paulo e Aspectos Biológicos dos Prin- tionship of Balistes showed significant differences cipais Recursos Pesqueiros Demersais Costeiros da Região between sexes. The exponent b of this relation Sudeste/Sul do Brasil (23°-29°S). Dr.Sc. thesis, Instituto Oceanográfico da Universidade de São Paulo, (Brasil). (3.265 - males and 3.340 - females) showed that Caverivière, A., M. Kulbicki, J. Konan and F. Gerlotto. – 1981. the growth is allometric. Bilan des connaissances actuelles sur Balistes carolinensis dans le Golfe de Guinée. Doc. Sci. Rech. Océanogr. Abidjan 12(1): 1-78. Ford, E. – 1933. An account of the investigations conduct- ed at Plymouth during the years from 1924-1933. J. Mar. Biol. ACKNOWLEDGEMENTS Assoc. U.K., 19: 305-384. Jonhson, A.G. and C.H. Saloman. – 1984. Age, growth, and mor- tality of gray triggerfish, Balistes capriscus, from the North- I would like to thank Drs. Noriyoshi Yamaguti eastern Gulf of Mexico. Fish. Bull. 82(3): 485-492. and Carmen Lúcia del Bianco Rossi-Wogtschowski, Ofori-Danson, P.K. – 1989. Growth of grey triggerfish Balistes capriscus based on growth checks of dorsal spine. Fishbyte. whose suggestions greatly improved this study, and December 1989. CAPES and FAPESP for the scholarship and finan- Taylor, C.C. – 1960. Temperature, growth and mortality - The pacific cockle. J. Cons. Int. Explor. Mer. 26: 117-124. cial support for this research. I am also grateful to Walford, L.A. –1946. A new graphic method of describing the both referees for their suggestions and valuable growth of . Biol. Bull. 90: 141-147. comments. Scient. ed.: M. Harmelin-Vivien

AGE, GROWTH AND LONGEVITY OF BALISTES CAPRISCUS 173